
import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt
from statsmodels.tsa.seasonal import seasonal_decompose

pd.set_option('display.max_rows', 1000)


# Load the data
df = pd.read_csv('/Users/mustafacan/Desktop/all_ticks_wide.csv')
df['timestamp'] = pd.to_datetime(df['timestamp'])


filtered_df = df[(df['timestamp'].dt.year >= 2014) & (df['timestamp'].dt.year <= 2017)]


# Use linear interpolation to fill N/A values
interpolated_data = filtered_df.set_index('timestamp').interpolate(method='time').reset_index()


# Extract year and month for grouping
interpolated_data['Year'] = interpolated_data['timestamp'].dt.year
interpolated_data['Month'] = interpolated_data['timestamp'].dt.month


na_counts = filtered_df.isna().sum()[1:].sort_values()
na_counts[na_counts <= 1000]

GARAN    398
THYAO    407
HALKB    428
KRDMD    434
YKBNK    435
VAKBN    438
TUPRS    450
AKBNK    452
EREGL    456
ISCTR    456
SAHOL    462
TKFEN    466
PETKM    466
TCELL    468
TTKOM    471
ASELS    473
PGSUS    475
SISE     477
GUBRF    478
VESTL    484
OTKAR    485
FROTO    488
KCHOL    491
ARCLK    503
TOASO    508
TRKCM    514
GOODY    518
CCOLA    540
BRISA    545
ZOREN    548
KARSN    576
MGROS    584
BAGFS    606
ALARK    638
SODA     708
AKSA     715
PRKME    720
TSKB     775
AKSEN    776
ANACM    801
ECILC    834
AEFES    928
dtype: int64


selected_companies = ['GARAN', 'AKBNK', 'YKBNK', 'THYAO', 'PGSUS', 'SAHOL', 'KCHOL']


fig, axes = plt.subplots(len(selected_companies), 4, figsize=(20, 20), constrained_layout=True)

for i, company in enumerate(selected_companies):
    # Finding the min and max for the specific company
    company_min = interpolated_data[company].min()
    company_max = interpolated_data[company].max()
    
    for j, year in enumerate(range(2014, 2018)):
        monthly_data = interpolated_data[interpolated_data['Year'] == year]
        sns.boxplot(x="Month", y=company, data=monthly_data, ax=axes[i, j])
        axes[i, j].set_title(f"{company} - {year}")
        axes[i, j].set_xlabel("")
        axes[i, j].set_ylim(company_min, company_max)
        if j == 0:
            axes[i, j].set_ylabel("Price")
        else:
            axes[i, j].set_ylabel("")


plt.show()


monthly_stats = interpolated_data.groupby(['Year', 'Month'])[selected_companies].agg(['mean', 'std'])

fig, axes = plt.subplots(len(selected_companies), 4, figsize=(20, 20), constrained_layout=True)

for i, company in enumerate(selected_companies):
    for j, year in enumerate(range(2014, 2018)):
        data_for_year = interpolated_data[interpolated_data['Year'] == year]
        
        # Extract monthly means and standard deviations for the specific company and year
        company_monthly_stats = monthly_stats[company].loc[year]
        
        # Compute the 3-sigma bounds for each month
        upper_bound = company_monthly_stats['mean'] + 3 * company_monthly_stats['std']
        lower_bound = company_monthly_stats['mean'] - 3 * company_monthly_stats['std']
        
        # Plot data points and the 3-sigma bounds
        sns.scatterplot(x='Month', y=company, data=data_for_year, ax=axes[i, j], color='blue', label='Data Points')
        axes[i, j].plot(company_monthly_stats.index, upper_bound, color='red', label='Mean + 3σ')
        axes[i, j].plot(company_monthly_stats.index, lower_bound, color='green', label='Mean - 3σ')
        
        axes[i, j].set_title(f"{company} - {year}")
        axes[i, j].set_xlabel("")
        if j == 0:
            axes[i, j].set_ylabel("Price")
        else:
            axes[i, j].set_ylabel("")
        if i == 0 and j == 0:
            axes[i, j].legend()


plt.show()


def plot_generator(year):
    fig, axes = plt.subplots(7, 12, figsize=(40, 20), constrained_layout=True)

    daily_stats = interpolated_data.groupby(['Year', 'Month', 'timestamp'])[selected_companies].mean()

    for j, company in enumerate(selected_companies):
        for x, year in enumerate(range(year, year+1)):
            for i, month in enumerate(range(1, 13)):


                data_for_year = data_for_month = daily_stats.xs((year))[company].reset_index()

                # Filter the data for the specific month across all years
                data_for_month = daily_stats.xs((year,month))[company].reset_index()

                # Compute the 3-sigma bounds for the month for the specific company
                yearly_std = data_for_year[company].std()
                yearly_mean = data_for_year[company].mean()

                monthly_mean = data_for_month[company].mean()
                monthly_std = data_for_month[company].std()

                upper_bound = monthly_mean + 3 * monthly_std
                lower_bound = monthly_mean - 3 * monthly_std

                yearly_upper_bound = yearly_mean + 3 * yearly_std
                yearly_lower_bound = yearly_mean - 3 * yearly_std


                # Plot data points and the 3-sigma bounds
                axes[j, i].plot(data_for_month['timestamp'].dt.day, data_for_month[company], color='blue', label=f"{company} Data Points")
                axes[j, i].axhline(upper_bound, color='red', linestyle='--', label=f"{company} Mean + 3σ")
                axes[j, i].axhline(lower_bound, color='green', linestyle='--', label=f"{company} Mean - 3σ")

                axes[j, i].set_title(f"{company} - Month {month} - Year {year}")
                axes[j, i].set_xlabel("Day of Month")
                axes[j, i].set_ylim(monthly_mean + 5 * monthly_std, monthly_mean - 5 * monthly_std)
                
    plt.show()

plot_generator(2014)
plot_generator(2015)
plot_generator(2016)
plot_generator(2017)


data = pd.read_csv('/Users/mustafacan/Desktop/all_ticks_wide.csv')

# Convert the 'timestamp' column to datetime format
data['timestamp'] = pd.to_datetime(data['timestamp'])

# Extract year, month, and day from the 'timestamp' column
data['year'] = data['timestamp'].dt.year
data['month'] = data['timestamp'].dt.month
data['day'] = data['timestamp'].dt.day
data['hour'] = data['timestamp'].dt.hour

# Limit the time interval from the beginning of 2014 to the end of 2017
data = data[(data['year'] >= 2014) & (data['year'] <= 2017)]

# List of company columns (excluding the timestamp and the extracted date columns)
companies = ['GARAN', 'AKBNK', 'YKBNK', 'THYAO', 'PGSUS', 'SAHOL', 'KCHOL']

outliers = []

# Iterate through each company
for company in companies:
    # Group by year and month, and calculate mean and standard deviation
    grouped = data.groupby(['year', 'month'])[company].agg(['mean', 'std'])
    
    # Merge the grouped data back to the original dataframe
    merged_data = data.merge(grouped, how='left', on=['year', 'month'])
    
    # Calculate the 3-sigma limits
    merged_data['lower_bound'] = merged_data['mean'] - 3 * merged_data['std']
    merged_data['upper_bound'] = merged_data['mean'] + 3 * merged_data['std']
    
    # Filter rows where the stock data is outside the 3-sigma limits
    outlier_data = merged_data[(merged_data[company] < merged_data['lower_bound']) | 
                               (merged_data[company] > merged_data['upper_bound'])]
    
    # Extract relevant information for the outliers
    for index, row in outlier_data.iterrows():
        outliers.append({
            'company': company,
            'year': row['year'],
            'month': row['month'],
            'day': row['day'],
            'hour': row['hour'],
            'stock_value': row[company]
        })

# Convert the list of outliers to a DataFrame
outliers_df = pd.DataFrame(outliers)

outliers_df


data = pd.read_csv('/Users/mustafacan/Desktop/all_ticks_wide.csv')

# List of companies to filter
companies = ['GARAN', 'AKBNK', 'YKBNK', 'THYAO', 'PGSUS', 'SAHOL', 'KCHOL']

# Convert the timestamp column to datetime type
data['timestamp'] = pd.to_datetime(data['timestamp'])

# Filter the data based on the given time range
filtered_data = data[
    (data['timestamp'] >= '2014-01-01') & 
    (data['timestamp'] <= '2017-12-31')
][['timestamp'] + companies]

# Extract year and month from the timestamp
filtered_data['year'] = filtered_data['timestamp'].dt.year
filtered_data['month'] = filtered_data['timestamp'].dt.month

grouped = filtered_data.groupby(['year', 'month'])

# Calculate the mean and standard deviation for each group
stats = grouped[companies].agg(['mean', 'std'])

# Calculate the UCL and LCL
for company in companies:
    stats[(company, 'UCL')] = stats[(company, 'mean')] + 3 * stats[(company, 'std')]
    stats[(company, 'LCL')] = stats[(company, 'mean')] - 3 * stats[(company, 'std')]


interpolated_data = filtered_data.copy()

# Apply linear interpolation to each column based on the timestamp
for company in companies:
    interpolated_data[company] = interpolated_data[company].interpolate(method='linear')


# Group by year and month, and then calculate the mean and standard deviation for each company's column
grouped_interpolated = interpolated_data.groupby(['year', 'month'])

# Calculate the mean and standard deviation for each group
stats_interpolated = grouped_interpolated[companies].agg(['mean', 'std'])

# Calculate the UCL and LCL
for company in companies:
    stats_interpolated[(company, 'UCL')] = stats_interpolated[(company, 'mean')] + 3 * stats_interpolated[(company, 'std')]
    stats_interpolated[(company, 'LCL')] = stats_interpolated[(company, 'mean')] - 3 * stats_interpolated[(company, 'std')]


outliers = []

# Iterate over each company to identify outliers
for company in companies:
    for (year, month), group in interpolated_data.groupby(['year', 'month']):
        UCL = stats_interpolated.loc[(year, month), (company, 'UCL')]
        LCL = stats_interpolated.loc[(year, month), (company, 'LCL')]
        
        # Find the outliers in the current group for the current company
        outlier_data = group[(group[company] > UCL) | (group[company] < LCL)]
        
        for idx, row in outlier_data.iterrows():
            # Extract the +/- 5 days surrounding the outlier day
            start_date = row['timestamp'] - pd.Timedelta(days=5)
            end_date = row['timestamp'] + pd.Timedelta(days=5)
            
            surrounding_data = interpolated_data[(interpolated_data['timestamp'] >= start_date) & 
                                                 (interpolated_data['timestamp'] <= end_date)]
            
            outliers.append({
                'company': company,
                'outlier_date': row['timestamp'],
                'surrounding_data': surrounding_data
            })

# Convert the list of outliers to a DataFrame for better visualization
outliers_df = pd.DataFrame(outliers)

# Displaying the number of outliers detected for each company
outliers_count = outliers_df['company'].value_counts()

# Function to compute the combined intervals for a given company
def compute_intervals(company_data):
    combined_intervals = []
    
    # Sort the outlier data by date
    company_data = company_data.sort_values(by="outlier_date")

    # Initialize the first interval
    start_interval = company_data.iloc[0]['surrounding_data']['timestamp'].min()
    end_interval = company_data.iloc[0]['surrounding_data']['timestamp'].max()

    for idx, row in company_data.iterrows():
        current_start = row['surrounding_data']['timestamp'].min()
        current_end = row['surrounding_data']['timestamp'].max()
        
        # If the current interval intersects with the previous one, combine them
        if current_start <= end_interval:
            end_interval = max(end_interval, current_end)
        else:
            # If not, store the previous interval and start a new one
            combined_intervals.append((start_interval, end_interval))
            start_interval = current_start
            end_interval = current_end

    # Add the last interval
    combined_intervals.append((start_interval, end_interval))
    
    return combined_intervals

all_intervals = {}
for company in companies:
    company_data = outliers_df[outliers_df['company'] == company]
    all_intervals[company] = compute_intervals(company_data)

verbal_intervals = {}
for company, intervals in all_intervals.items():
    formatted_intervals = []
    for start, end in intervals:
        start_str = start.strftime('%dth of %B %Y')
        end_str = end.strftime('%dth of %B %Y')
        formatted_intervals.append((start_str, end_str))
    verbal_intervals[company] = formatted_intervals

verbal_intervals

{'GARAN': [('24th of November 2014', '03th of December 2014'),
  ('11th of December 2014', '22th of December 2014'),
  ('29th of January 2015', '06th of February 2015'),
  ('09th of December 2015', '18th of December 2015'),
  ('27th of April 2016', '06th of May 2016'),
  ('24th of February 2017', '06th of March 2017')],
 'AKBNK': [('11th of December 2014', '22th of December 2014'),
  ('03th of June 2015', '12th of June 2015'),
  ('27th of October 2016', '04th of November 2016'),
  ('26th of January 2017', '06th of February 2017'),
  ('01th of March 2017', '10th of March 2017'),
  ('04th of October 2017', '13th of October 2017')],
 'YKBNK': [('03th of June 2015', '12th of June 2015'),
  ('09th of December 2015', '18th of December 2015'),
  ('28th of March 2016', '05th of April 2016'),
  ('27th of October 2016', '04th of November 2016'),
  ('25th of January 2017', '03th of February 2017')],
 'THYAO': [('24th of November 2014', '03th of December 2014'),
  ('19th of September 2016', '28th of September 2016'),
  ('25th of November 2016', '05th of December 2016'),
  ('23th of February 2017', '03th of March 2017')],
 'PGSUS': [('27th of January 2014', '07th of February 2014'),
  ('05th of June 2014', '13th of June 2014'),
  ('24th of November 2014', '03th of December 2014'),
  ('03th of June 2015', '12th of June 2015'),
  ('26th of December 2016', '04th of January 2017'),
  ('23th of February 2017', '03th of March 2017')],
 'SAHOL': [('06th of June 2014', '16th of June 2014'),
  ('24th of November 2014', '03th of December 2014'),
  ('30th of December 2015', '08th of January 2016'),
  ('25th of January 2016', '03th of February 2016'),
  ('27th of April 2016', '06th of May 2016'),
  ('22th of May 2017', '29th of May 2017'),
  ('04th of October 2017', '13th of October 2017')],
 'KCHOL': [('30th of January 2014', '07th of February 2014'),
  ('26th of May 2014', '13th of June 2014'),
  ('04th of October 2017', '13th of October 2017')]}


data = pd.read_csv('/Users/mustafacan/Desktop/all_ticks_wide.csv')
data['timestamp'] = pd.to_datetime(data['timestamp'])

# Define the list of selected companies and filter the data
selected_companies = ['GARAN', 'AKBNK', 'YKBNK', 'THYAO', 'PGSUS', 'SAHOL', 'KCHOL']
filtered_data = data[['timestamp'] + selected_companies]
filtered_data.set_index('timestamp', inplace=True)

# Filter the data for the date range from the beginning of 2014 to the end of 2017
start_date = '2014-01-01'
end_date = '2017-12-31'
filtered_data = filtered_data[(filtered_data.index >= start_date) & (filtered_data.index <= end_date)]

# Resample the data by month and compute the mean for each month
monthly_data = filtered_data.resample('M').mean()

# Function to perform decomposition and plotting for actual and trend for two-year cycles
def plot_decomposition_two_years(data, company, start_year, end_year):
    for cycle_start in range(start_year, end_year, 2):  # Step by 2 years for each cycle
        cycle_end = cycle_start + 1
        two_year_data = data[f'{cycle_start}':f'{cycle_end}']
        decomposition = seasonal_decompose(two_year_data[company], model='additive')

        # Plot actual data
        plt.figure(figsize=(14, 5))
        plt.subplot(1, 2, 1)
        plt.plot(decomposition.observed)
        plt.title(f'Actual Data for {company} ({cycle_start}-{cycle_end})')
        plt.xlabel('Month')
        plt.ylabel('Price')

        # Plot trend data
        plt.subplot(1, 2, 2)
        plt.plot(decomposition.trend)
        plt.title(f'Trend for {company} ({cycle_start}-{cycle_end})')
        plt.xlabel('Month')
        plt.ylabel('Price')

        plt.tight_layout()
        plt.show()

# Perform decomposition and plot for each company and two-year cycles
companies = ['GARAN', 'AKBNK', 'YKBNK', 'THYAO', 'PGSUS', 'SAHOL', 'KCHOL']
for company in companies:
    plot_decomposition_two_years(monthly_data, company, 2014, 2017)


# Load the data
data = pd.read_csv('/Users/mustafacan/Desktop/all_ticks_wide.csv')

# Convert the 'timestamp' column to datetime format
data['timestamp'] = pd.to_datetime(data['timestamp'])

# Extract year, month, and day from the 'timestamp' column
data['year'] = data['timestamp'].dt.year
data['month'] = data['timestamp'].dt.month
data['day'] = data['timestamp'].dt.day
data['hour'] = data['timestamp'].dt.hour

# Limit the time interval from the beginning of 2014 to the end of 2017
data = data[(data['year'] >= 2014) & (data['year'] <= 2017)]

# List of company columns (excluding the timestamp and the extracted date columns)
companies = ['GARAN', 'AKBNK', 'YKBNK', 'THYAO', 'PGSUS', 'SAHOL', 'KCHOL']

outliers = []

# Iterate through each company
for company in companies:
    # Group by year and month, and calculate mean and standard deviation
    grouped = data.groupby(['year', 'month'])[company].agg(['mean', 'std'])
    
    # Merge the grouped data back to the original dataframe
    merged_data = data.merge(grouped, how='left', on=['year', 'month'])
    
    # Calculate the 3-sigma limits
    merged_data['lower_bound'] = merged_data['mean'] - 3 * merged_data['std']
    merged_data['upper_bound'] = merged_data['mean'] + 3 * merged_data['std']
    
    # Filter rows where the stock data is outside the 3-sigma limits
    outlier_data = merged_data[(merged_data[company] < merged_data['lower_bound']) | 
                               (merged_data[company] > merged_data['upper_bound'])]
    
    # Extract relevant information for the outliers
    for index, row in outlier_data.iterrows():
        outliers.append({
            'company': company,
            'year': row['year'],
            'month': row['month'],
            'day': row['day'],
        })

# Convert the list of outliers to a DataFrame
outliers_df = pd.DataFrame(outliers)
outliers_df = outliers_df.drop_duplicates()

def generate_outlier_dates(stock:str, df):
    lst = []
    for _, row in df.iterrows():
        if row[0] == stock:
            lst.append(f"{row[1]}-{row[2]}-{row[3]}")
            
    return lst

def plot_yearly_stock_prices(df, stock_names, outlier_dates, start_year, end_year):
    # Group the data by year and stock name within the specified range
    for year in range(start_year, end_year + 1):
        if year in df['timestamp'].dt.year.unique():
            year_data = df[df['timestamp'].dt.year == year]
            for stock in stock_names:
                if stock in year_data.columns:
                    stock_data = year_data[['timestamp', stock]]

                    # Create a separate plot for each stock and year
                    plt.figure(figsize=(10, 5))  # Set the figure size

                    # Plot the stock data
                    plt.plot(stock_data['timestamp'], stock_data[stock], label=stock)

                    # Convert outlier dates from strings to datetime objects
                    outlier_dates_stock = [pd.to_datetime(date) for date in outlier_dates[stock_names.index(stock)]]

                    # Add vertical lines at the positions of outlier dates for the current stock
                    for date in outlier_dates_stock:
                        if date.year == year:
                            plt.axvline(date, color='red', linestyle='--', label=date.strftime('%Y-%m-%d'))

                    # Customize the plot
                    plt.title(f'Stock Prices for Year {year} - {stock}')
                    plt.xlabel('Date')
                    plt.ylabel('Stock Price')
                    plt.legend(loc='upper left')

                    # Save or display the plot
                    plt.savefig(f'yearly_stock_prices_{year}_{stock}.png')
                    plt.show()

list = [generate_outlier_dates("GARAN", outliers_df), generate_outlier_dates("AKBNK", outliers_df),
       generate_outlier_dates("YKBNK", outliers_df), generate_outlier_dates("THYAO", outliers_df),
       generate_outlier_dates("PGSUS", outliers_df), generate_outlier_dates("SAHOL", outliers_df),
       generate_outlier_dates("KCHOL", outliers_df)]
                    
plot_yearly_stock_prices(df, ['GARAN', 'AKBNK', 'YKBNK', 'THYAO', 'PGSUS', 'SAHOL', 'KCHOL'], list, 2014, 2017)

	company	year	month	day	hour	stock_value
0	GARAN	2014	11	28	14	8.8857
1	GARAN	2014	11	28	14	8.8766
2	GARAN	2014	12	16	14	7.5352
3	GARAN	2014	12	16	14	7.4271
4	GARAN	2014	12	16	15	7.4631
5	GARAN	2014	12	16	15	7.4722
6	GARAN	2014	12	16	15	7.4722
7	GARAN	2014	12	17	7	7.4722
8	GARAN	2014	12	17	7	7.5262
9	GARAN	2014	12	17	8	7.5711
10	GARAN	2014	12	17	8	7.5711
11	GARAN	2014	12	17	9	7.5711
12	GARAN	2014	12	17	9	7.5711
13	GARAN	2014	12	17	9	7.5622
14	GARAN	2014	12	17	10	7.5622
15	GARAN	2014	12	17	10	7.5531
16	GARAN	2014	12	17	10	7.5531
17	GARAN	2014	12	17	12	7.5622
18	GARAN	2015	2	2	9	9.3628
19	GARAN	2015	2	2	12	9.3628
20	GARAN	2015	2	2	12	9.3628
21	GARAN	2015	2	2	12	9.3628
22	GARAN	2015	2	2	12	9.3628
23	GARAN	2015	2	2	13	9.3628
24	GARAN	2015	2	2	14	9.3628
25	GARAN	2015	2	2	14	9.3628
26	GARAN	2015	2	2	15	9.3628
27	GARAN	2015	2	2	15	9.3628
28	GARAN	2015	2	3	7	9.4078
29	GARAN	2015	2	3	7	9.4078
30	GARAN	2015	12	14	14	6.2248
31	GARAN	2015	12	14	15	6.2156
32	GARAN	2015	12	14	15	6.2156
33	GARAN	2016	5	2	6	7.9700
34	GARAN	2016	5	2	6	7.9607
35	GARAN	2016	5	2	7	7.9515
36	GARAN	2016	5	2	7	7.9329
37	GARAN	2016	5	2	7	7.9422
38	GARAN	2016	5	2	8	7.9515
39	GARAN	2016	5	2	8	7.9329
40	GARAN	2016	5	2	8	7.9329
41	GARAN	2017	3	1	6	8.0444
42	GARAN	2017	3	1	7	8.0351
43	GARAN	2017	3	1	7	8.0351
44	GARAN	2017	3	1	7	8.0536
45	AKBNK	2014	12	16	14	6.0394
46	AKBNK	2014	12	16	15	6.0470
47	AKBNK	2014	12	16	15	6.0394
48	AKBNK	2014	12	17	7	6.0470
49	AKBNK	2015	2	3	7	7.1220
50	AKBNK	2015	6	8	6	5.5905
51	AKBNK	2015	6	8	6	5.6457
52	AKBNK	2015	6	8	7	5.6692
53	AKBNK	2015	6	8	7	5.6692
54	AKBNK	2015	6	8	8	5.6457
55	AKBNK	2015	6	8	8	5.6378
56	AKBNK	2015	6	8	8	5.6300
57	AKBNK	2015	6	8	8	5.6300
58	AKBNK	2015	6	8	9	5.6378
59	AKBNK	2015	6	8	9	5.6300
60	AKBNK	2015	6	8	11	5.6457
61	AKBNK	2016	11	1	6	6.6761
62	AKBNK	2016	11	1	7	6.6761
63	AKBNK	2016	11	1	8	6.6761
64	AKBNK	2016	11	1	11	6.6761
65	AKBNK	2016	11	1	11	6.6761
66	AKBNK	2017	1	31	11	6.7643
67	AKBNK	2017	1	31	14	6.7643
68	AKBNK	2017	1	31	14	6.7803
69	AKBNK	2017	2	1	8	6.6681
70	AKBNK	2017	2	1	8	6.6521
71	AKBNK	2017	3	6	12	7.4375
72	AKBNK	2017	3	6	13	7.4375
73	AKBNK	2017	3	6	13	7.4375
74	AKBNK	2017	3	6	13	7.4375
75	AKBNK	2017	3	6	13	7.4294
76	AKBNK	2017	3	6	14	7.4294
77	AKBNK	2017	3	7	6	7.4535
78	AKBNK	2017	3	7	7	7.4294
79	AKBNK	2017	3	7	7	7.4615
80	AKBNK	2017	3	7	7	7.4375
81	AKBNK	2017	3	7	7	7.4455
82	AKBNK	2017	10	9	7	7.3946
83	AKBNK	2017	10	9	7	7.4357
84	AKBNK	2017	10	9	7	7.4439
85	AKBNK	2017	10	9	8	7.4357
86	AKBNK	2017	10	9	8	7.4192
87	AKBNK	2017	10	9	8	7.4275
88	AKBNK	2017	10	9	8	7.4192
89	AKBNK	2017	10	9	9	7.4275
90	AKBNK	2017	10	9	9	7.4357
91	AKBNK	2017	10	9	9	7.4604
92	YKBNK	2014	5	30	6	3.0947
93	YKBNK	2014	5	30	6	3.0947
94	YKBNK	2014	5	30	11	3.0947
95	YKBNK	2014	5	30	14	3.0947
96	YKBNK	2015	6	8	11	2.3713
97	YKBNK	2015	12	14	12	2.0685
98	YKBNK	2015	12	14	13	2.0685
99	YKBNK	2015	12	14	13	2.0685
100	YKBNK	2015	12	14	14	2.0685
101	YKBNK	2015	12	14	14	2.0685
102	YKBNK	2015	12	14	15	2.0685
103	YKBNK	2015	12	14	15	2.0685
104	YKBNK	2015	12	15	7	2.0685
105	YKBNK	2016	3	31	12	2.7129
106	YKBNK	2016	11	1	6	2.4036
107	YKBNK	2016	11	1	7	2.4036
108	YKBNK	2016	11	1	8	2.4036
109	YKBNK	2017	1	30	14	2.5324
110	YKBNK	2017	1	30	15	2.5453
111	YKBNK	2017	1	31	6	2.5324
112	YKBNK	2017	1	31	7	2.5324
113	YKBNK	2017	1	31	7	2.5260
114	YKBNK	2017	1	31	7	2.5260
115	YKBNK	2017	1	31	7	2.5067
116	YKBNK	2017	1	31	8	2.4873
117	YKBNK	2017	1	31	9	2.4873
118	YKBNK	2017	1	31	9	2.4873
119	YKBNK	2017	1	31	9	2.4873
120	YKBNK	2017	1	31	9	2.4873
121	YKBNK	2017	1	31	10	2.4873
122	YKBNK	2017	1	31	11	2.4938
123	YKBNK	2017	1	31	11	2.5002
124	YKBNK	2017	1	31	11	2.5002
125	YKBNK	2017	1	31	11	2.4873
126	YKBNK	2017	1	31	12	2.4873
127	THYAO	2014	11	28	9	9.4000
128	THYAO	2016	9	23	9	5.3900
129	THYAO	2016	9	23	9	5.3900
130	THYAO	2016	9	23	9	5.3900
131	THYAO	2016	9	23	10	5.3900
132	THYAO	2016	9	23	11	5.4000
133	THYAO	2017	2	28	12	5.4600
134	THYAO	2017	2	28	12	5.4500
135	THYAO	2017	2	28	12	5.4200
136	THYAO	2017	2	28	13	5.4200
137	THYAO	2017	2	28	13	5.4400
138	THYAO	2017	2	28	13	5.4600
139	THYAO	2017	2	28	13	5.4500
140	THYAO	2017	2	28	14	5.4600
141	THYAO	2017	2	28	14	5.4600
142	THYAO	2017	2	28	14	5.4500
143	THYAO	2017	2	28	14	5.4500
144	THYAO	2017	2	28	15	5.4600
145	PGSUS	2014	1	31	14	26.5000
146	PGSUS	2014	1	31	14	26.4000
147	PGSUS	2014	1	31	15	26.2000
148	PGSUS	2014	1	31	15	26.0000
149	PGSUS	2014	1	31	15	26.0000
150	PGSUS	2014	2	3	7	24.9000
151	PGSUS	2014	6	10	7	30.5000
152	PGSUS	2014	6	10	7	30.4500
153	PGSUS	2014	6	10	7	30.4000
154	PGSUS	2014	6	10	8	30.5500
155	PGSUS	2014	6	10	8	30.4500
156	PGSUS	2014	6	10	8	30.5000
157	PGSUS	2014	6	10	11	30.4500
158	PGSUS	2014	6	10	12	30.4000
159	PGSUS	2014	11	28	9	33.5500
160	PGSUS	2014	11	28	9	33.3500
161	PGSUS	2014	11	28	9	33.3000
162	PGSUS	2014	11	28	9	33.4000
163	PGSUS	2014	11	28	10	33.0000
164	PGSUS	2014	11	28	10	33.0500
165	PGSUS	2014	11	28	12	33.0000
166	PGSUS	2014	11	28	12	33.0000
167	PGSUS	2014	11	28	12	33.1500
168	PGSUS	2014	11	28	12	33.2000
169	PGSUS	2014	11	28	13	33.1500
170	PGSUS	2014	11	28	13	33.1500
171	PGSUS	2014	11	28	13	33.1500
172	PGSUS	2014	11	28	13	33.1500
173	PGSUS	2014	11	28	14	33.0500
174	PGSUS	2014	11	28	14	33.0500
175	PGSUS	2014	11	28	14	33.1000
176	PGSUS	2014	11	28	14	33.0000
177	PGSUS	2015	6	8	6	22.9500
178	PGSUS	2016	12	30	14	13.9400
179	PGSUS	2016	12	30	14	14.2700
180	PGSUS	2016	12	30	14	14.2600
181	PGSUS	2016	12	30	14	14.2100
182	PGSUS	2016	12	30	15	14.1500
183	PGSUS	2017	2	28	7	15.2500
184	PGSUS	2017	2	28	8	15.2600
185	PGSUS	2017	2	28	8	15.2500
186	PGSUS	2017	2	28	8	15.2300
187	PGSUS	2017	2	28	8	15.2400
188	PGSUS	2017	2	28	9	15.2600
189	PGSUS	2017	2	28	9	15.2400
190	PGSUS	2017	2	28	9	15.2600
191	PGSUS	2017	2	28	11	15.2600
192	PGSUS	2017	2	28	11	15.2600
193	PGSUS	2017	2	28	11	15.2400
194	PGSUS	2017	2	28	11	15.2300
195	PGSUS	2017	2	28	12	15.2200
196	PGSUS	2017	2	28	12	15.1600
197	PGSUS	2017	2	28	12	15.1000
198	PGSUS	2017	2	28	12	15.0600
199	PGSUS	2017	2	28	13	15.1200
200	PGSUS	2017	2	28	13	15.1600
201	PGSUS	2017	2	28	13	15.1900
202	PGSUS	2017	2	28	13	15.2100
203	PGSUS	2017	2	28	14	15.1800
204	PGSUS	2017	2	28	14	15.1800
205	PGSUS	2017	2	28	14	15.1400
206	PGSUS	2017	2	28	14	15.1500
207	PGSUS	2017	2	28	15	15.1600
208	SAHOL	2014	6	11	6	9.3639
209	SAHOL	2014	6	11	6	9.3639
210	SAHOL	2014	11	28	14	9.6315
211	SAHOL	2014	12	1	12	9.8099
212	SAHOL	2014	12	1	12	9.8099
213	SAHOL	2016	1	4	7	6.6797
214	SAHOL	2016	1	29	12	7.6712
215	SAHOL	2016	1	29	12	7.6802
216	SAHOL	2016	1	29	14	7.6712
217	SAHOL	2016	1	29	14	7.6802
218	SAHOL	2016	1	29	14	7.6712
219	SAHOL	2016	1	29	15	7.6712
220	SAHOL	2016	1	29	15	7.6802
221	SAHOL	2016	1	29	15	7.7072
222	SAHOL	2016	5	2	6	9.2469
223	SAHOL	2016	5	2	6	9.2561
224	SAHOL	2016	5	2	7	9.2103
225	SAHOL	2016	5	2	7	9.2194
226	SAHOL	2016	5	2	7	9.2286
227	SAHOL	2016	5	2	8	9.2378
228	SAHOL	2016	5	2	8	9.2103
229	SAHOL	2016	5	2	8	9.2286
230	SAHOL	2016	5	2	9	9.2103
231	SAHOL	2016	5	2	12	9.2103
232	SAHOL	2016	5	2	13	9.2103
233	SAHOL	2017	5	24	14	10.3698
234	SAHOL	2017	5	24	15	10.3698
235	SAHOL	2017	10	9	6	9.0619
236	SAHOL	2017	10	9	7	8.9217
237	SAHOL	2017	10	9	7	8.9592
238	SAHOL	2017	10	9	7	8.9685
239	SAHOL	2017	10	9	7	8.9685
240	SAHOL	2017	10	9	8	8.9872
241	SAHOL	2017	10	9	8	8.9685
242	SAHOL	2017	10	9	8	8.9778
243	SAHOL	2017	10	9	8	8.9872
244	SAHOL	2017	10	9	9	8.9685
245	SAHOL	2017	10	9	9	8.9872
246	KCHOL	2014	2	4	7	6.8699
247	KCHOL	2014	2	4	8	6.8879
248	KCHOL	2014	2	4	9	6.8879
249	KCHOL	2014	5	30	6	9.4213
250	KCHOL	2014	5	30	11	9.4213
251	KCHOL	2014	5	30	11	9.5115
252	KCHOL	2014	5	30	12	9.4663
253	KCHOL	2014	5	30	12	9.4663
254	KCHOL	2014	5	30	12	9.4663
255	KCHOL	2014	5	30	12	9.5115
256	KCHOL	2014	5	30	13	9.4663
257	KCHOL	2014	5	30	13	9.4213
258	KCHOL	2014	5	30	13	9.4213
259	KCHOL	2014	6	9	11	9.8720
260	KCHOL	2014	6	9	12	9.8720
261	KCHOL	2017	10	9	7	14.7493
262	KCHOL	2017	10	9	7	14.8162
263	KCHOL	2017	10	9	7	14.8352
264	KCHOL	2017	10	9	7	14.7779
265	KCHOL	2017	10	9	8	14.8066
266	KCHOL	2017	10	9	8	14.7588
267	KCHOL	2017	10	9	8	14.7970
268	KCHOL	2017	10	9	8	14.7970
269	KCHOL	2017	10	9	9	14.7875
270	KCHOL	2017	10	9	9	14.8162
271	KCHOL	2017	10	9	11	14.7970
272	KCHOL	2017	10	9	11	14.8256
273	KCHOL	2017	10	9	11	14.8066
274	KCHOL	2017	10	9	11	14.7970
275	KCHOL	2017	10	9	12	14.7588
276	KCHOL	2017	10	9	12	14.7779
277	KCHOL	2017	10	9	12	14.7779
278	KCHOL	2017	10	9	12	14.7970
279	KCHOL	2017	10	9	13	14.8066

IE 423 - FALL 2023 - PROJECT HOMEWORK PART 1¶

Bengisu Hacet¶

Mustafa Can Coşkun¶

Fethi Kahvecioğlu¶

BOX PLOT ANALYSIS¶

Yearly 3-Sigma Chart with Scattered Data Points over Y-Axis¶

3 SIGMA CHART¶

Listed all the outlier points to make internal check of our analysis.¶

OUTLIER INTERVAL LISTING¶

Trend Analysis¶

OUTLIER DAYS VISUALIZED IN THE PLOT¶

Market Dynamics and News Related to Fluctuations of Stocks¶

Airline Industry¶

Pegasus¶

2014¶

Search results for stock names¶

3rd June 2015 - 12nd June 2015¶

Search names for stock names¶

26th December 2016 - 4th January 2017¶

Turkish Airlines¶

24th November 2014 - 03rd December 2014¶

19th September 2016 - 28th September 2016¶

25th November 2016 - 05th December 2016¶

23rd February 2017 - 03rd March 2017¶

Financial Services¶

Garanti¶

24 November 2014 - 03 December 2014 & 11 December 2014 & 22 December 2014¶

09 December 2015 - 18 December 2015¶

27 April 2016 - 06 May 2016¶

24 February 2017 - 06 March 2017¶

YapıKredi¶

03 June 2015 - 12 June 2015¶

09 December 2015 - 18 December 2015¶

28 March 2016 - 05 April 2016¶

27 October 2016 - 04 November 2016¶

25 January 2017 - 03 February 2017¶

Akbank¶

03 June 2015 - 12 June 2015¶

27 October 2016 - 04 November 2016¶

26 January 2017 - 06 February 2017 & 01 March 2017 - 10 March 2017¶

04 October 2017 - 13 October 2017¶

Conglomerate¶

Koc Holding¶

30 January 2014 - 07 February 2014¶

26 May 2014 - 13 June 2014¶

04 October 2017 - 13 October 2017:¶

Sabanci Holding¶

06 June 2014 - 16 June 2014:¶

24 November 2014 - 03 December 2014:¶

30 December 2015 - 08 January 2016¶

25 January 2016 - 03 February 2016¶

27 April 2016 - 06 May 2016¶

22 May 2017 - 29 May 2017¶